Presentation Authors: Zoe Gan, Case Wood*, Allison Deal, Chang Xu, Yue Wang, Angela Smith, Raj Pruthi, Chapel Hill, NC
Introduction: Work relative value units (RVUs) represent a metric of physician productivity, but the correlation between RVUs and surgical complexity remains poorly understood. Our primary objective was to define the correlation of RVUs with measures of surgical complexity and physician workload in the field of urology. We hypothesized that RVUs would correlate poorly with these measures. Our secondary objectives were to 1) identify procedures with outlying RVU values for their measures of surgical complexity and workload and 2) calculate projected RVU values for these procedures.
Methods: We retrospectively reviewed the 2012-2016 American College of Surgeons National Surgical Quality Improvement Program (NSQIP) databases, from which we selected 56 current procedural terminology (CPT) codes representing the spectrum of urologic surgery. Linear regression was used to correlate RVUs with length of hospital stay (LOS), operating room (OR) time, morbidity, mortality, serious adverse events (SAEs), and readmissions. Variables with R2>0.7 were incorporated into a multivariable model. Studentized residuals were used to identify outlying procedures. For outlying procedures, projected RVU values were estimated based on the regression model.
Results: We identified 190,323 urologic cases over a five-year period for the 56 selected procedures. RVUs correlated well with LOS (R2 = 0.80), OR time (R2 = 0.87), and morbidity (R2 = 0.74). RVUs only moderately correlated with SAEs (R2 = 0.64), mortality (R2 = 0.52), and readmissions (R2 = 0.57). Outlying procedures identified using the multivariable regression model were retroperitoneal lymph node dissection (projected +10.64 RVUs), laparoscopic partial nephrectomy (projected -6.83 RVUs), laparoscopic ureteroneocystotomy (projected -6.78 RVUs), and cystectomy with bilateral pelvic lymphadenectomy (projected +6.54 RVUs).
Conclusions: In urology, certain measures of surgical complexity and physician workload appear to correlate with RVUs more than others. Incorporating objective data may improve RVU assignments for individual CPT codes in the future.